mirrors/postgres
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Update float dpow() comment about whick platforms had issues with Nan.

Browse Source 
Stefan Kaltenbrunner
This commit is contained in:
Bruce Momjian 2007-01-05 03:19:26 +00:00
parent 53c26d6f8f
commit 138668ca37
1 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
src/backend/utils/adt/float.c
View File

@ -8,7 +8,7 @@
* *
* *
* IDENTIFICATION * IDENTIFICATION
* $PostgreSQL: pgsql/src/backend/utils/adt/float.c,v 1.140 2007/01/04 05:18:39 momjian Exp $ * $PostgreSQL: pgsql/src/backend/utils/adt/float.c,v 1.141 2007/01/05 03:19:26 momjian Exp $
* *
*------------------------------------------------------------------------- *-------------------------------------------------------------------------
*/ */
@ -1443,8 +1443,8 @@ dpow(PG_FUNCTION_ARGS)
* follows _IEEE_, _POSIX_, _XOPEN_, or _SVID_, so we try to avoid * follows _IEEE_, _POSIX_, _XOPEN_, or _SVID_, so we try to avoid
* using errno. However, some platform/CPU combinations return * using errno. However, some platform/CPU combinations return
* errno == EDOM and result == Nan, so we have to check for that and * errno == EDOM and result == Nan, so we have to check for that and
* set result properly. For example, Linux on Pentium, pre-Xeon * set result properly. For example, Linux on 32-bit x86 hardware
* hardware returns EDOM/Nan for (-1) ^ 1e19, but (-1) ^ 1e18 retuns * returns EDOM/Nan for (-1) ^ 1e19, but (-1) ^ 1e18 returns
* 1 -- basically a negative base raised to a very high power causes * 1 -- basically a negative base raised to a very high power causes
* it on some CPUs. * it on some CPUs.
*/ */